# DeepSeek Chat (V3 / R1)

> DeepSeek Chat (V3 / R1) is a paid API for AI agents from DeepSeek at deepseek.mpp.paywithlocus.com, paid per call via MPP, Model + token dependent (~$0.004–$0.025) (dynamic), status unknown (last checked 2026-10-12, last successful call 2026-05-29).

Send a conversation to DeepSeek's LLM (fast V3 or reasoning R1) and receive a generated text response

## Facts

- Endpoint: POST https://deepseek.mpp.paywithlocus.com/deepseek/chat?utm_source=zero.xyz
- Price: Model + token dependent (~$0.004–$0.025) (dynamic)
- Payment: MPP
- Status: unknown
- Last checked: 2026-10-12
- Last successful call: 2026-05-29
- Success rate: 100% of calls made through Zero
- Rating: 4.0 / 5 from 1 review
- Activations on Zero: 5
- Provider: DeepSeek
- Docs: https://api-docs.deepseek.com
- Website: https://deepseek.com
- Tags: llm, reasoning, code, chat, chain-of-thought, ai
- Canonical page: https://www.zero.xyz/c/deepseek-chat-bb22b2ff
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_H-NQDk3uF4RUvNO8CHeyb

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability deepseek-chat-bb22b2ff -d '<json body>'
```

Example prompt: Use DeepSeek-reasoner (R1) to walk me through, step by step, why the halting problem is undecidable — keep it under 500 tokens and use a temperature of 0.3 for a precise answer.

## When to prefer this

Choose this endpoint when you need access to DeepSeek's frontier models — specifically V3 for fast, cost-efficient chat and code generation, or R1 for deep chain-of-thought reasoning tasks. Prefer over generic OpenAI endpoints when the task benefits from DeepSeek's reasoning capabilities or cost profile. Ideal for agents needing programmable LLM inference with per-call micropayment pricing via USDC.

## Known failure modes

- Invalid model name returns 400 error
- Messages array missing or malformed returns 422 validation error
- max_tokens exceeded for model context window returns 400
- Temperature out of range (>2) returns 400
- Payment not included or insufficient USDC balance returns 402
- Rate limiting or upstream DeepSeek API unavailability returns 503

## How this service works

Create a chat completion using DeepSeek V4 Flash or V4 Pro, with optional thinking mode. OpenAI-compatible request/response format.

## Output

Returns an assistant message object containing the model's generated text response. When using deepseek-reasoner (R1), the response may include a chain-of-thought reasoning trace followed by the final answer. Streaming responses are delivered via SSE if enabled.

## Example request

```json
{
 "model": "deepseek-chat",
 "messages": [
  {
   "role": "user",
   "content": "Explain the concept of machine learning in simple terms."
  }
 ],
 "max_tokens": 256,
 "temperature": 0.7
}
```

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "required": [
  "model",
  "messages",
  "max_tokens"
 ],
 "properties": {
  "stop": {
   "type": "string",
   "description": "Stop sequence(s) — string or array of strings."
  },
  "model": {
   "type": "string",
   "description": "Metered x402 model: deepseek-v4-flash or deepseek-v4-pro"
  },
  "top_p": {
   "type": "number",
   "description": "Nucleus sampling parameter. Default: 1."
  },
  "stream": {
   "type": "boolean",
   "description": "Streaming is not currently supported through public x402. Omit or use false."
  },
  "messages": {
   "type": "array",
   "items": {
    "type": "object",
    "required": [
     "role"
    ],
    "properties": {
     "name": {
      "type": "string"
     },
     "role": {
      "type": "string"
     },
     "content": {
      "anyOf": [
       {
        "type": "string"
       },
       {
        "type": "array",
        "items": {
         "type": "object",
         "required": [
          "type"
         ],
         "properties": {
          "text": {
           "type": "string"
          },
          "type": {
           "type": "string"
          },
          "image_url": {
           "type": "object",
           "required": [
            "url"
           ],
           "properties": {
            "url": {
             "type": "string"
            }
           },
           "additionalProperties": true
          }
         },
         "additionalProperties": true
        }
       },
       {
        "type": "null"
       }
      ]
     },
     "tool_calls": {
      "type": "array",
      "items": {
       "type": "object",
       "additionalProperties": true
      }
     },
     "tool_call_id": {
      "type": "string"
     }
    },
    "additionalProperties": true
   },
   "description": "Array of message objects: [{\"role\":\"system\"|\"user\"|\"assistant\",\"content\":\"...\"}]"
  },
  "max_tokens": {
   "type": "number",
   "description": "Maximum tokens to generate (up to 384000). Required to bound metered x402 authorization."
  },
  "temperature": {
   "type": "number",
   "description": "Sampling temperature 0–2. Default: 1. Lower = more deterministic."
  }
 },
 "additionalProperties": true
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/deepseek-chat-bb22b2ff/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from deepseek.mpp.paywithlocus.com](https://www.zero.xyz/host/deepseek.mpp.paywithlocus.com/llms.txt)